Treatment for stimulant addiction is often complicated by relapse. Behavioral factors contributing to relapse include cue-induced drug craving and an inability to weigh the long-term consequences of a choice. These states have been linked, respectively, to limbic activation and to defects in ventral orbitofrontal brain function. Functional brain imaging provides a noninvasive means of studying alterations in regional brain function accompanying behavioral states in humans, and the advent of functional magnetic resonance imaging (fMRI) has greatly increased it's accessibility. While the majority of fMRI studies have used blood oxygenation level dependent (BOLD) contrast as a marker for regional neural activation, this approach also has shortcomings, including lack of absolute quantification, baseline drift with poor sensitivity for very low frequency events, and signal loss in certain brain regions such as the ventral orbitofrontal cortex due to static susceptibility effects. Over the past decade, we have been developing methods for direct measurement of cerebral blood flow (CBF) using MRI. This class of techniques, termed arterial spin labeled (ASL) perfusion MRI, utilizes magnetically labeled arterial blood water as an endogenous tracer. Our preliminary data demonstrate that: 1) CBF values measured using ASL perfusion MRI are reproducible, 2) task-activation measured by ASL shows less intersubject variability than BOLD, 3) ASL contrast shows noise characteristics that make it suitable for studying sequential changes in resting or activated brain function over long periods, and 4) signal changes detected by ASL can be measured in the presence of large static susceptibility gradients. We propose to further develop and validate ASL perfusion MRI methods for the purpose of enhancing our investigations into the neurobiology of cocaine addiction. Specifically, we will improve the measurement of CBF in regions of high static susceptibility (e.g, the amygdala and ventral frontal lobes), implement ASL perfusion MRI at high field, and validate the use of ASL for examining CBF under resting and activated conditions in cocaine patients and contrcontrols. The anticipated findings of limbic activation during cue-induced craving, resting limbic hypoperfusion, and ventral orbitofrontal perfusion deflects may reflect core vulnerabilites for relapse and will provide direct evidence for the advantages of ASL perfusion in the study of addiction-related states.